The rise of AI and tts nexus with energy

Artificial intelligence has moved rapidly from a niche technology to a major economic force, with significant implications for the energy system. AI relies heavily on electricity, particularly across data centres, networks and hardware manufacturing. At the same time, AI can enhance how energy is produced, transported and consumed. The report frames this two-way relationship as central to future energy security, affordability and sustainability.

AI capabilities are improving quickly, especially in large language models, driving higher computational intensity. This raises concerns about electricity demand growth, system reliability and emissions, while also creating opportunities for optimisation and efficiency across the energy value chain.

Energy for AI

Electricity demand from data centres is rising sharply due to AI workloads. In the base case, global data centre electricity consumption more than doubles by 2030 compared with 2022 levels. AI-optimised servers consume significantly more power than conventional computing, increasing peak load pressures on grids.

The report highlights risks of grid connection delays, local congestion and higher system costs, particularly where data centres cluster geographically. Procurement strategies such as long-term power purchase agreements and on-site generation are increasingly used by technology firms. Operational and locational flexibility of data centres, including demand response, is identified as critical to reducing system stress and supporting grid stability.

AI for energy optimisation

AI applications can materially reduce energy demand and operating costs across the energy system. In industry, AI-driven process optimisation, predictive maintenance and digital twins improve productivity while lowering energy use. Optimisation of production processes delivers the largest direct energy savings.

In transport, AI-enabled eco-driving can reduce fuel consumption by 2–10%, while autonomous trucks demonstrate fuel savings of 10–20%. Under widespread adoption, AI could reduce road freight energy demand by over 1.5 exajoules by 2035, around 4% of total freight demand.

In buildings and power systems, AI improves forecasting, load balancing and asset management, enhancing system resilience and reliability. However, the report notes rebound risks, where efficiency gains may be offset by increased demand if not managed through policy and system design.

AI for energy innovation

AI accelerates innovation across energy technologies by shortening research cycles and reducing costs. In batteries, catalysts, carbon capture materials and cement production, AI supports material discovery, performance modelling and process design.

For carbon capture, AI has enabled the rapid screening of thousands of metal–organic frameworks, significantly accelerating laboratory research. However, deployment remains constrained by limited real-world operational data and scale-up challenges. In cement, AI can help reduce clinker content, lowering both energy use and emissions, but progress is limited by low R&D investment, historically below 1% of sector revenues.

Emerging themes on energy and AI

AI introduces new energy security risks linked to critical minerals, hardware supply chains and cyber resilience. Smart integration of data centres is required to avoid exacerbating system vulnerabilities.

Digital and AI skills are emerging as a bottleneck. While demand for digital skills in energy sectors rose by around 20% between 2018 and 2023, AI-specific hiring lags other industries. Entry-level AI roles pay around 30% less in energy than in technology, limiting talent attraction. Upskilling, reskilling and partnerships with external providers are identified as necessary responses.

In emerging market and developing economies, limited electricity reliability, internet access and high data costs constrain AI deployment. Only around 60% of populations in these economies have reliable internet access, and households spend significantly more on broadband than in advanced economies. Coordinated investment in energy and digital infrastructure is required to unlock inclusive AI benefits.

The AI and energy policy landscape

Governments play a key role in enabling beneficial AI deployment through data governance, infrastructure investment and regulatory clarity. Policies that align AI deployment with emissions reduction, system resilience and affordability are essential.

The report concludes that AI can either increase or reduce emissions depending on how electricity supply, efficiency gains and rebound effects are managed. Strategic policy frameworks and system-level planning are therefore critical to ensuring AI supports, rather than undermines, energy transition objectives.